6 
Transactions Texas Academy op Science. 
[38] 
was very much smaller. In one sample from El Paso, selected because 
it showed an abnormal amount of sediment, I found something over fif- 
teen per cent, of silt at the end of ope week, but this was chosen as repre- 
senting the extreme case. 
However, for the Brazos and Wichita rivers the extreme percentages 
of silt are away below this. For the Brazos river the largest percentage 
of silt for any one set of samples was less than four, while for the Wich- 
ita river the maximum percentage was found on July 19, 1900, to be 
6.84. This was on a rising river in the early stages of a small flood. In 
fact the smaller floods generally show a higher percentage of silt than 
do the larger ones, and there is more sediment when the stream is rising 
than after it has attained its highest stage. For such determinations as 
I have myself made the results were obtained at the end of a week and 
should be reduced about one-fourth in order to obtain the percentages at 
the end of a year, as has already been noted. 
For the Wichita River there are only two general conditions of the 
water. When the river is low the water is clear, or nearly so; when the 
river is up the water is red, sometimes intensely so. For the Brazos River 
there are three conditions of the water. First, when the stream is low 
it is clear or nearly so, though when looked at from above both streams 
may appear red on account of the red sediment at the bottom of the 
stream. When the flood in the Brazos comes from near its head the water 
is red and very similar to the water in the Wichita at high stages, because 
the same character of country constitutes the watersheds. When the rains 
that fall over the black lands, and Cretaceous areas generally, swell the 
Brazos the water is dark and the percentage of silt small, while the red 
rises carry the largest percentages. 
In order to determine the effect that the silt in any stream would have 
on an impounding reservoir it is necessary to know also the discharge of 
the stream at the time the collections were made. Indeed, both discharge 
measurements and silt determination should be made daily if very close 
results are desired. On both the Brazos and the Wichita rivers I have 
established gauging stations at which the gauge heights are read daily. 
These heights, together with as many discharge measurements as could 
be made, enabled me to compute the daily discharge, and as the observer 
also noted the appearance of the water daily it was possible for me to 
approximate the mean monthly discharge of water and silt for the two 
streams during the periods covered. Doing this for the Brazos river, for 
a period covering fifteen months, I found, as the approximate percentage 
of silt at the end of one week, 1.28, which if reduced to the probable per- 
centage at the end of one year’s settlement would yield 0.96 per cent. 
Taking only the latter twelve months of this period I got practically the 
same results. However, the last two years have yielded discharges much 
higher than the average, because the floods have been- more frequent, so 
